Wireless Communications Trends and Challenges ANDREA J GOLDSMITH

Wireless Communications: Trends and Challenges ANDREA J. GOLDSMITH Dept. of Electrical Engineering Stanford University http: //ee. stanford. edu/~andrea 8 C 32810. 1 -Cimini-7/98

OUTLINE • Introduction • Radio Environment • Physical Layer Issues • Channel Access Issues • Network Issues • Standards and Future Systems • Summary 8 C 32810. 20 -Cimini-7/98

WIRELESS DATA VISION Region TAXI City laptops, PDAs Campus In-Building Seamless Multimedia Networks with Mobility and Freedom from Tethers [R. Katz, "Does Wireless Data Have a Future? ", Plenary Talk, INFOCOM '96] 8 C 32810. 35 -Cimini-7/98

VOICE VERSUS DATA VERSUS VIDEO Voice Data Video < 100 ms – < 100 ms < 1% 0 <1% BER 10 -2 - 10 -3 < 10 -5 < 10 -7 Data Rate 8 -32 kbps 1 -100 Mbps 1 -20 Mbps Continuous Bursty Continuous Delay Packet Loss Traffic Wired Networks Trying to Integrate (ATM, SONET, Multimedia Services) 8 C 32810. 80 -Cimini-7/98

WHAT IS THE FUTURE OF WIRELESS DATA? 100 USA market 90 80 millions 70 cellular + PCS subs Internet users 60 50 paging subs 40 30 dedicated wireless data subs 20 10 annual laptop sales 0 1995 *Estimates as of 1996 7 C 29822. 008 -Cimini-9/97 laptop users 2000

THE ISSUE IS PERFORMANCE "The mobile data market has been slow to take off, but progress is being made. The most formidable obstacle to user acceptance remains performance. " I. Brodsky, "Countdown to Mobile Blast Off", Network World, February 19, 1996 Mobile Multimedia Terminal Network Radio Adaptation Protocols & Control & Modem Wireless Interface Radio Link Mobility Control Protocols Network Interface Radio Port RADIO ACCESS SEGMENT Signaling/ Routing Mobility Control MOBILE NETWORK SEGMENT BROADBAND WIRELESS NETWORK • Link Performance: Data Rate and Quality • Network Performance: Access, Coverage, Reliability, Qo. S, and Internetworking 8 C 32810. 81 -Cimini-7/98

GAP BETWEEN WIRED AND WIRELESS NETWORK CAPABILITIES LOCAL AREA PACKET SWITCHING 100 M 100, 000 WIDE AREA CIRCUIT SWITCHING ATM 100, 000 Ethernet 10, 000 FDDI wired- wireless Ethernet bit-rate "gap" 100 (kbps) 1 st gen WLAN Polling 2 nd gen WLA N 8 C 32810. 79 -Cimini-7/98 (kbps) 28. 8 modem 9. 6 modem 2. 4 modem Packet 1 Radio 32 kbps PCS 14. 4 9. 6 cellular digital cellular 2. 4 cellular . 1 . 01 1970 100 ISDN bit-rate "gap" 10 10 1 wired- wireless 1000 User Bit-Rate . 01 1980 YEAR 1990 2000 1970 1980 YEAR 1990 2000

TECHNICAL CHALLENGES • Low-Power/Low-Cost Implementations • Scarce Radio Spectrum • Radio Channel Characteristics – Limits on Signal Coverage – Limits on Data Rates • Efficient Network Architectures and Protocols • Seamless Internetworking • Authentication and Security 7 C 29822. 010 -Cimini-9/97

RADIO ENVIRONMENT • Path Loss • Shadow Fading • Multipath 8 C 32810. 83 -Cimini-7/98 Limit the Bit Rate and/or Coverage

PATH LOSS MODEL • Different, often complicated, models are used for different environments. • A simple model for path loss, L, is L= Pr Pt =K 1 f da 2 where Pr is the local mean received signal power, Pt is the transmitted power, d is the transmitter-receiver distance, f is frequency, and K is a transmission constant. The path loss exponent a = 2 in free space; 2 £ a £ 4 in typical environments. 7 C 29822. 011 -Cimini-9/97

SHADOW FADING • The received signal is shadowed by obstructions such as hills and buildings. • This results in variations in the local mean received signal power, Pr (d. B) = Pr (d. B) + Gs where Gs ~ N(0, ss ), 4 £ ss £ 10 d. B. 2 • Implications – nonuniform coverage – increases the required transmit power R 8 C 32810. 84 -Cimini-7/98 P = P r 0

MULTIPATH Received Power t Delay Spread h(t) = S aiejq d(t-ti) i i • Constructive and Destructive Interference of Arriving Rays 10 0 d. B With Respect -10 to RMS Value -20 -30 0. 5 l 8 C 32810. 85 -Cimini-7/98 0 0. 5 0 10 1 t, in seconds 20 x, in wavelength 1. 5 30

Received Power DELAY SPREAD TIME DOMAIN INTERPRETATION Two-ray model t = rms delay spread 2 t Delay Channel Input 1 0 T 1 T Channel Output t small 0 T 2 T 2 T t large T t • T small • t large negligible intersymbol interference significant intersymbol interference, T which causes an irreducible error floor 8 C 32810. 88 -Cimini-7/98

PHYSICAL LAYER ISSUES • Link Performance Measures • Modulation Tradeoffs • Flat Fading Countermeasures • Delay Spread Countermeasures 8 C 32810. 91 -Cimini-7/98

LINK PERFORMANCE MEASURES PROBABILITY OF BIT ERROR • The probability of bit error, Pb, in a radio environment is a random variable. – average Pb, Pb – Pr [Pb > Pbtarget] D outage, Pout = • Typically one of these measures is useful, depending on the Doppler frequency and the bit rate. 8 C 32810. 92 -Cimini-7/98

LINK PERFORMANCE MEASURES EFFICIENCY • Spectral Efficiency – a measure of the data rate per unit bandwidth for a given bit error probability and transmitted power • Power Efficiency – a measure of the required received power to achieve a given data rate for a given bit error probability and bandwidth • Throughput/Delay 8 C 32810. 72 -Cimini-7/98

GOALS OF MODULATION TECHNIQUES • High Bit Rate • High Spectral Efficiency • High Power Efficiency • Low-Cost/Low-Power Implementation • Robustness to Impairments 8 C 32810. 74 -Cimini-7/98

DIGITAL MODULATION • Any modulated signal can be represented as s(t) = A(t) cos [wct + f(t)] amplitude phase or frequency s(t) = A(t) cos f(t) cos wct in-phase - A(t) sin f(t) sin wct quadrature • Linear versus nonlinear modulation Þ impact on spectral efficiency • Constant envelope versus non-constant envelope Þ hardware implications with impact on power efficiency 8 C 32810. 94 -Cimini-7/98

LINEAR MODULATION TECHNIQUES s(t) = [ S an g (t-n. T)]cos wct - [ S bn g (t-n. T)] sin wct n n I(t), in-phase Q(t), quadrature LINEAR MODULATIONS Square Constellations M-ARY QUADRATURE AMPLITUDE MOD. (M-QAM) M¹ 4 M=4 (4 -QAM = 4 -PSK) CONVENTIONAL 4 -PSK (QPSK) 8 C 32810. 95 -Cimini-7/98 M-ARY PHASE SHIFT KEYING (M-PSK) Circular Constellations M¹ 4 OFFSET DIFFERENTIAL 4 -PSK (OQPSK) (DQPSK, p/4 -DQPSK)

SIGNAL CONSTELLATIONS M-PSK (Circular Constellations) bn 4 -PSK 16 -PSK an M-QAM (Square Constellations) bn 16 -QAM 4 -PSK an Tradeoffs – Higher-order modulations (M large) are more spectrally efficient but less power efficient. – M-QAM is more spectrally efficient than M-PSK but also more sensitive to system nonlinearities. 8 C 32810. 96 -Cimini-7/98

PULSE SHAPING Rectangular pulses are spectrally inefficient pulse shaping intersymbol interference (ISI) non-constant envelope Nyquist pulses 8 C 32810. 75 -Cimini-7/98

RAISED COSINE PULSE SHAPING g(t) a = 0, 0. 5 a=1 0 T 2 T 3 T a = 0. 5 a=0 a = 0. 5 G(f) a=1 1 2 T Relative Peak Instantaneous Power (d. B) - 1 T 1 2 T 1 T 30 256 QAM 64 QAM 20 16 PSK 16 QAM 10 8 PSK 4 PSK 0 8 C 32810. 97 -Cimini-7/98 0 0. 25 0. 75 1. 0 Cosine Rolloff Factor, a 4 T f t

DEMODULATION • Coherent detection requires a coherent phase reference. – difficult to obtain in a rapidly fading environment – increases receiver complexity • Differential detection uses the previous symbol for the reference signal. – eliminates need for coherent reference – entails loss in power efficiency (up to 3 d. B) – Doppler causes irreducible error floor, typically small for high bit rates 8 C 32810. 133 -Cimini-7/98

FREQUENCY SHIFT KEYING • Continuous Phase FSK (CPFSK) – digital data encoded in the frequency shift – typically implemented with frequency modulator to maintain continuous phase t s(t) = A cos [wct + 2 pkf ò d(t) dt] -¥ – nonlinear modulation but constant-envelope • Minimum Shift Keying (MSK) – minimum bandwidth, sidelobes large – can be implemented using I-Q receiver • Gaussian Minimum Shift Keying (GMSK) – reduces sidelobes of MSK using a premodulation filter – used by RAM Mobile Data, CDPD, and HIPERLAN 8 C 32810. 98 -Cimini-7/98

SPECTRAL CHARACTERISTICS 10 QPSK/DQPSK GMSK Power Spectral Density (d. B) 0 -20 -40 -60 B 3 -d. BTb = 0. 16 0. 25 -80 1. 0 -100 -120 (MSK) 0 0. 5 1. 0 1. 5 2. 0 Normalized Frequency (f-fc)Tb 7 C 29822. 013 -Cimini-9/97 2. 5

BIT ERROR PROBABILITY AWGN CHANNEL 10 -1 5 2 10 -2 5 For Pb = 10 -3 2 Pb BPSK 6. 5 d. B QPSK 6. 5 d. B DBPSK ~8 d. B DQPSK ~9 d. B DBPSK 10 -3 BPSK, QPSK 5 2 DQPSK 10 -4 5 2 10 -5 5 2 10 -6 0 2 4 6 8 gb, SNR/bit, d. B 10 12 14 • QPSK is more spectrally efficient than BPSK with the same performance. • M-PSK, for M>4, is more spectrally efficient but requires more SNR per bit. • There is ~3 d. B power penalty for differential detection. 8 C 32810. 99 -Cimini-7/98

BIT ERROR PROBABILITY FADING CHANNEL 1 5 2 10 -2 5 Pb DBPSK 2 BPSK 10 -3 5 AWGN 2 10 -4 5 2 10 -5 0 5 10 15 20 25 30 35 gb, SNR/bit, d. B • Pb is inversely proportion to the average SNR per bit. • Transmission in a fading environment requires about 18 d. B more power for Pb = 10 -3. 8 C 32810. 100 -Cimini-7/98

BIT ERROR PROBABILITY EFFECTS OF DOPPLER SPREAD • Doppler causes an irreducible error floor when differential detection is used Þ decorrelation of reference signal. 100 QPSK DQPSK 10 -1 Rayleigh Fading 10 -2 -3 Pb 10 f. DT=0. 003 No Fading 10 -4 0. 002 0. 001 10 -5 0 10 -6 0 10 20 30 40 50 60 gb, SNR/bit, d. B • The irreducible Pb depends on the data rate and the Doppler. For f. D = 80 Hz, data rate T Pbfloor 10 kbps 10 -4 s 3 x 10 -4 100 kbps 10 -5 s 3 x 10 -6 1 Mbps 10 -6 s 3 x 10 -8 The implication is that Doppler is not an issue for high-speed wireless data. [M. D. Yacoub, Foundations of Mobile Radio Engineering , CRC Press, 1993] 8 C 32810. 101 -Cimini-7/98

BIT ERROR PROBABILITY EFFECTS OF DELAY SPREAD • ISI causes an irreducible error floor. Irreducible Pb 10 -1 Coherent Detection + BPSK QPSK OQPSK Modulation x MSK x 10 -2 x x + + x + 10 -3 + x + 10 -4 10 -2 • 10 -1 100 rms delay spread t = symbol period T The rms delay spread imposes a limit on the maximum bit rate in a multipath environment. For example, for QPSK, t Maximum Bit Rate Mobile (rural) 25 msec 8 kbps Mobile (city) 2. 5 msec 80 kbps Microcells 500 nsec 400 kbps Large Building 100 nsec 2 Mbps [J. C. -I. Chuang, "The Effects of Time Delay Spread on Portable Radio Communications Channels with Digital Modulation, " IEEE JSAC, June 1987] 8 C 32810. 102 -Cimini-7/98

SUMMARY OF MODULATION ISSUES • Tradeoffs – linear versus nonlinear modulation – constant envelope versus non-constant envelope – coherent versus differential detection – power efficiency versus spectral efficiency • Limitations – flat fading – doppler – delay spread 8 C 32810. 103 -Cimini-7/98

HOW DO WE OVERCOME THE LIMITATIONS IMPOSED BY THE RADIO CHANNEL? • Flat Fading Countermeasures – Fade Margin – Diversity – Coding and Interleaving – Adaptive Techniques • Delay Spread Countermeasures – Equalization – Multicarrier – Spread Spectrum – Antenna Solutions 8 C 32810. 104 -Cimini-7/98

DIVERSITY • Independent signal paths have a low probability Received Signal Power (d. Bm) of experiencing deep fades simultaneously. 0 -20 -40 -60 -80 -100 0 4 8 12 16 d The chance that two deep fades occur simultaneously is rare. • The basic concept is to send the same information over independently fading radio • Independent fading paths can be achieved by separating the signal in time, frequency, space, polarization, etc. 8 C 32810. 105 -Cimini-7/98

DIVERSITY COMBINING TECHNIQUES • • • a 1 a 2 a 3 a. M Combiner Output • Selection Combining: picks the branch with the highest SNR. • Equal-Gain Combining: all branches are coherently combined with equal weights. • Maximal-Ratio Combining: all branches are coherently combined with weights which depend on the branch SNR. 8 C 32810. 106 -Cimini-7/98

DIVERSITY PERFORMANCE • There is dramatic improvement even with two-branch selection combining. – 10 d. B reduction in required SNR for 1% outage Þ less transmitted power or higher bit rates or larger coverage area Pb 10 -1 5 Pout 99. 99 99. 5 98. 0 90. 0 80. 0 70. 0 60. 0 50. 0 40. 0 30. 0 Maximal Ratio Combining 2 10 -2 5 M=2 20. 0 2 10. 0 M=1 5. 0 2. 0 10 -4 5 1. 0 0. 5 M=2 2 1 2 Maximal Ratio Equal Gain = 5 M 10 -3 Selection 0. 2 0. 1 10 -5 5 0. 05 M=4 2 0. 02 10 -6 5 10 15 20 25 30 gb, SNR/bit, d. B 35 40 0. 01 -40 -30 -20 10 log ( -10 1 margin 0 ) • The output SNR with Maximal-Ratio Combining improves linearly with the number of diversity branches, M complexity becomes prohibitive. 7 C 29822. 014 -Cimini-9/97 Þ the 10

CHANNEL CODING • Channel coding reduces Pb by introducing redundancy in the transmitted bit stream. • Block and convolutional codes acheive this improvement at the expense of increased signal bandwidth or a lower data rate. • Bit error probability–AWGN channel 10 -2 5 BPSK 2 10 -3 5 For Pb = 10 -6 2 Uncoded Hamming BCH Conv. 10 -4 5 Pb Uncoded 2 Hamming (7, 4, 1) 10 -5 10. 5 d. B 10. 0 d. B 6. 5 d. B 5. 0 d. B BCH (127, 64, 10) 5 Conv. 1/2 rate (k=7) 2 10 -6 5 2 10 -7 0 2 4 6 8 10 12 14 gb, SNR/bit, d. B • Fading causes burst errors. If the fading is slow enough relative to the symbol rate, coding will not be effective. 7 C 29822. 015 -Cimini-9/97

CODING PERFORMANCE FADING CHANNEL • Pb performance for the IS-136 rate-1/2 convolutional code on a simulated mobile radio channel (harddecision decoding). 1 Uncoded 50 km/hr Coded 1 km/hr Coded 8 km/hr Coded 50 km/hr Coded 100 km/hr 10 -1 Pb 10 -2 10 -3 10 -4 8 10 12 14 16 gb, SNR/bit, d. B 18 20 • Negligible coding gain if fading is slow compared to bit rate Þ interleaving [V. Iyengar and J. Michaelides, "Performance Evaluations of RLPs (Radio Link Protocols) for TDMA Data Services, " ITIA Contribution TR 45. 3. 2. 5/93. 03. 30. 10, Chicago, March 30, 1993] 8 C 32810. 17 -Cimini-7/98

CODING PERFORMANCE FADING CHANNEL • Pb performance for the IS-136 rate-1/2 convolutional code on a simulated mobile radio channel (soft decision decoding). 1 Uncoded 50 km/hr Coded 1 km/hr Coded 8 km/hr Coded 50 km/hr Coded 100 km/hr -1 10 10 -2 Pb 10 -3 10 -4 10 -5 8 10 12 14 16 gb, SNR/bit, d. B 18 20 [V. Iyengar and J. Michaelides, "Performance Evaluations of RLPs (Radio Link Protocols) for TDMA Data Services, " ITIA Contribution TR 45. 3. 2. 5/93. 03. 30. 10, Chicago, March 30, 1993] 8 C 32810. 18 -Cimini-7/98

CODING AND INTERLEAVING • The basic principle is to spread the burst errors over many code words. 1 codeword read into interleaver by rows 1 2 3 4 5 6 7 8 9 10 11 12 read out by columns 1, 5, 9, 2, 6, 10, 3, 7, 11, 4, 8, 12 channel reads out by rows 1, 2 , 3, 4, 5, 6 , 7, 8, 9, 10 , 11, 12 1 2 3 4 5 6 7 8 9 10 11 12 1, 5, 9, 2 , 6 , 10 , 3, 7, 11, 4, 8, 12 reads in by rows • The required interleaver size can be large if the relative fading rate is slow, as is usually the case for high-speed data. For example, f. D = 10 Hz, bit rate = 10 Mb/s, error burst = 330, 000 bits. delay and complexity 8 C 32819. 16 -Cimini-7/98

ADVANCED CODING TECHNIQUES • Trellis Codes – reduce Pb without bandwidth expansion through joint design of the channel code and signal constellation – can be designed with “built-in” time diversity • Turbo Codes – exhibit enormous coding gains – interleaving inherent to code design – very complex with large delays – not well-understood for fading channels 8 C 32810. 19 -Cimini-7/98

CODING PERFORMANCE TCM 10 8 PSK TCM -1 10 -3 10 -4 Uncoded 4 PSK Ungerboeck Code R=2/3, M=4 Pb 10 -2 10 -5 10 -6 MSB b 3, R=2/3 b 1, R=2/3 b 2, R=2/3 10 12 14 16 18 Es/N 0 (d. B) 8 C 32810. 69 -Cimini-7/98 LSB 20 22

ADAPTIVE TECHNIQUES • Adaptive Modulation • Automatic Repeat Request 8 C 32810. 21 -Cimini-7/98

ADAPTIVE MODULATION TRANSMITTER RECEIVER noise Adaptive Modulation and Coding Power Control Channel + Demodulation and Decoding Channel Estimate Delay FEEDBACK CHANNEL • Power and/or data rate adapted at transmitter to channel conditions • Potential for large increase in spectral efficiency • Can be combined with adaptive compression – requires reliable feedback channel and accurate channel estimation – increases transmitter and receiver complexity 8 C 32810. 22 -Cimini-7/98

AUTOMATIC REPEAT REQUEST (ARQ) • Method of "self-adapting" the data rate to the channel conditions • Used in combination with error-detecting code • Variations of ARQ used in Mobitex and CDPD • Types: Stop-and-Wait, Go-Back-N, Selective. Repeat – power and spectrally inefficient – impacts higher layer protocols – necessary for meeting stringent Pb requirements or data 7 C 29822. 023 -Cimini-9/97

DELAY SPREAD COUNTERMEASURES • Signal Processing – at the receiver, to alleviate the problems caused by delay spread (equalization) – at the transmitter, to make the signal less sensitive to delay spread (multicarrier, spread spectrum) • Antenna Solutions – change the environment to reduce, or eliminate, the delay spread (distributed antenna system, small cells, directive antennas) 7 C 29822. 024 -Cimini-9/97

EQUALIZER TYPES AND STRUCTURES The goal of equalization is to cancel the ISI or, equivalently, to flatten the frequency response. Equalizer Types Linear Nonlinear DFE ML Symbol Detector MLSE Structures Transversal Lattice Transversal Channel Estimator [J. G. Proakis, "Adaptive Equalization for TDMA Digital Mobile Radio, " IEEE Trans. on Veh. Tech. , May 1991] 8 C 32810. 107 -Cimini-7/98

LINEAR EQUALIZER • A linear equalizer effectively inverts the channel. n(t) Channel Hc(f) Equalizer 1 Heq(f)» Hc(f) • The linear equalizer is usually implemented as a tapped delay line. • On a channel with deep spectral nulls, this equalizer enhances the noise. poor performance on frequency-selective fading channels 8 C 32810. 108 -Cimini-7/98

DECISION FEEDBACK EQUALIZER DFE n(t) x(t) Hc(f) Forward Filter ^ x(t) + Feedback Filter • The DFE determines the ISI from the previously detected symbols and subtracts it from the incoming symbols. • This equalizer does not suffer from noise enhancement because it estimates the channel rather than inverting it. Þ The DFE has better performance than the linear equalizer in a frequency-selective fading channel. • The DFE is subject to error propagation if decisions are made incorrectly. • Decisions are made on coded symbols. Þ no coding gain 7 C 29822. 025 -Cimini-9/97

MAXIMUM LIKELIHOOD SEQUENCE ESTIMATION • MLSE has theoretically optimum performance. • It requires knowledge of the channel parameters and the noise distribution. • The implementation complexity grows exponentially with the length of the channel impulse response Þ not practical for high bit rates. 8 C 32810. 109 -Cimini-7/98

EQUALIZER ISSUES FOR HIGH-SPEED WIRELESS DATA • The number of required equalizer taps, N, is proportional to the delay spread. • The equalizer taps must be adapted at the highest Doppler rate. – The length and periodicity of the training sequence impacts the spectral efficiency. – There is a tradeoff between speed of convergence and complexity. Algorithms (for DFE) Least Mean Square (LMS) Number of Multiply Operations 2 N + 1 Convergence ~10 -100 N Advantages Disadvantages Low computational Slow complexity depends on convergence, channel Kalman 2. 5 N 2 + 4. 5 N Recursive Least Squares (RLS) ~N Fast convergence, High good tracking ability computational complexity Square Root 1. 5 N 2 + 6. 5 N ~N Better stability than Kalman High computational complexity Fast Kalman 20 N + 5 ~N Fast convergence and good tracking Could be unstable 7 C 29822. 026 -Cimini-9/97

EQUALIZER PERFORMANCE BPSK 1 10 Mbps 5 10 -1 10 -2 1 10 Pb 10 -3 10 -4 10 -5 . 1 5. 1 1 no equalizer DFE 10 -6 25 30 35 40 45 50 SNR (d. B) BPSK 1 16 Mbps 8 4, 16 10 -1 1 8. 1. 1, 4 1 Pout 10 -2 10 -3 no equalizer DFE 10 -4 1 10 -4 10 -8 Target Pb 10 -12 • Pahlavan has shown that, for 30 -meter cells (t = 50 ns), 20 Mb/s can be achieved using a DFE with 3 forward taps and 3 feedback taps. [K. Pahlavan, S. J. Howard, and T. A. Sexton, "Decision Feedback Equalization of the Indoor Radio Channel, " IEEE Trans. on Commun. , January 1993] 8 C 32810. 110 -Cimini-7/98

MULTICARRIER MODULATION • The transmission bandwidth is divided into many narrow subchannels which are transmitted in parallel. • Ideally, each subchannel is narrow enough so that the fading it experiences is flat Þ no ISI. Transmitter R/N b/s QAM filter d 0(t) f 0 d 1(t) D(t) f 1 d N-1(t) f. N-1 Bandlimited signals f 0 Receiver f 2 f 1 filter f 0 RF filter f 1 filter f. N-1 8 C 32810. 111 -Cimini-7/98 QAM f 0 QAM f 1 QAM f. N-1 RF

OFDM RECEIVER STRUCTURE Receiver ò f 0 RF ò d(0) d(1) f 1 ò parallel to serial converter QAM d(N-1) f N-1 • Subchannel Separation 1 NT ^ – integrate over NT, then d(m) = d(m) – choose fn = f 0 + n. Df, with Df = • Efficient FFT Implementation • A guard interval can virtually eliminate ISI (or, interblock interference) Þ lower spectral or power efficiency. 8 C 32810. 113 -Cimini-7/98

WHAT TO DO WITH BAD SUBCHANNELS? • Coding Across Subchannels Þ works best with large delay spread • Frequency Equalization Þ requires accurate channel estimation • Adaptive Loading Þ requires reliable feedback channel and accurate channel estimation 8 C 32810. 114 -Cimini-7/98

MULTICARRIER MODULATION ISSUES FOR HIGH-SPEED WIRELESS DATA • Minimal training is required. • Time-varying fading, frequency offset, and timing mismatch impair the orthogonality of the subchannels. • Large peak-to-average power ratio is a serious problem when transmitting through a nonlinearity. – possible solutions: nonlinear coding, clipping and filtering 8 C 32810. 115 -Cimini-7/98

CURRENT AND PROPOSED APPLICATIONS OF OFDM • Asymmetric Digital Subscriber Line • Digital Audio Broadcasting • Wireless LAN • Digital Terrestrial Television • High Speed Cellular 8 C 32810. 116 -Cimini-7/98

SPREAD SPECTRUM • Spread spectrum increases the transmit signal bandwidth to reduce the effects of flat fading, ISI and interference. • SS is used in all wireless LAN products in the ISM band – required for operation with reasonable power levels – minimal performance impact on other systems – IEEE 802. 11 standard • There are two SS methods: direct sequence and frequency hopping. – Direct sequence multiplies the data sequence by a faster chip sequence. – Frequency hopping varies the carrier frequency by the same chip sequence. 8 C 32810. 117 -Cimini-7/98

DIRECT SEQUENCE SPREAD SPECTRUM Interference Data (T) Modulator Carrier Recovery Channel Spreading (PN) Code Tc << T Demod Spreading (PN) Code Synch Transmitter Receiver Narrowband Filter Narrowband Interference Data Signal with Spreading Modulated Data 8 C 32810. 117 -Cimini-7/98 Data (T) Original Data Signal ISI Receiver Input Other SS Users Demodulator Filtering ISI

RAKE RECEIVER sc(t) Received Signal sc(t-Tc) Coherent Combiner Data Output Demodulator sc(t-2 Tc) • • • sc(t-TM) • When the chip time is much less than the rms delay spread, each branch has independent fading Þ equivalent to diversity combining. • When the chip time is greater than the rms delay spread, the paths cannot be resolved Þ no diversity gain. 8 C 32810. 119 -Cimini-7/98

PERFORMANCE OF RAKE RECEIVER FADING CHANNEL 0. 5 DPSK 10 -1 Rayleigh 10 -2 Pb 10 -3 RAKE AWGN 10 -4 10 -5 8 C 32810. 27 -Cimini-7/98 0 5 10 gb, SNR/bit, d. B 15

SPREAD SPECTRUM ISSUES FOR HIGH-SPEED WIRELESS DATA • Hardware Complexity – synchronization – high processing speeds for high bit rates – RAKE receiver • High Required Bandwidth to Accommodate Spreading Spread spectrum is difficult at high bit rates and not really needed. 8 C 32810. 120 -Cimini-7/98

ANTENNA SOLUTIONS Goal: Reduce (or eliminate) delay spread • Distributed Antenna System • Very Small Cells Þ antenna in every room • Sectorization • Directive Antennas/Beam Steering Omnidirectional 120 90 60 150 180 0 330 210 7 C 29822. 028 -Cimini-9/97 120 30 240 270 Sectorized 300 90 Directive 60 150 120 30 180 0 330 210 240 270 300 90 60 150 30 180 0 330 210 240 270 300

DISTRIBUTED ANTENNA SYSTEM Probability Abscissa Exceeded 1 Distributed Monopoles. 5 0 Central Monopole 0 10 20 30 RMS Delay Spread (ns) 40 50 [A. A. M. Saleh, A. J. Rustako, Jr. , and R. S. Roman, "Distributed Antennas for Indoor Radio Communications, " IEEE Trans. on Commun. , December 1987] 7 C 29822. 029 -Cimini-9/97

EXAMPLES OF PERFORMANCE IMPROVEMENTS • High-Speed Narrowbeam Antenna Experiment [P. F. Driessen "Gigabit/s Indoor Wireless Systems with Directional Antennas, " IEEE Trans. on Comm. , August 1996] – directional antennas (15° beamwidth) at both ends of LOS link – no equalization – 622 Mbps BPSK transmission without errors • Sectored Antennas [G. Yang and K. Pahlavan, "Comparative Performance Evaluation of Sector Antenna and DFE Systems in Indoor Radio Channels, " Proc. of ICC '92] – 6 sectors at base and mobile – best combination chosen – for Pout = 0. 01, 5 Mbps with omni, 25 Mbps with sectored antenna Omnidirectional Antennas Six Sector Antennas 1. 1 . 001 Pout 1 Mbps -3 -4 -5 -6 -7 - -9 -10 -11 -12 8 C 32810. 121 -Cimini-7/98 . 005 20 Mbps . 00 2. 001 10 Mbps . 0005 -13 10 10 -110 -210 10 10 8 Target Pb 30 Mbps . 01 10 Mbps 5 Mbps 3 Mbps 2 Mbps Pout. 01 . 0001 0 . 02 30 Mbps 20 Mbps 100 10 -1 -210 -310 -410 -510 -610 -710 -10 -910 -1010 -1110 -1210 -13 10 8 Target Pb

SUMMARY OF COUNTERMEASURES • Diversity • Coding and Interleaving • Adaptive Techniques • Equalization • Multicarrier • Spread Spectrum • Antenna Solutions These techniques can be combined. 8 C 32810. 123 -Cimini-7/98

COMBINED EQUALIZATION AND SECTORED ANTENNAS 1 Pt = 100 m. W Rb = 20 Mbps Omni . 1 Pout Omni+DFE . 01 Sector Sec+DFE . 001. 0001 20 40 60 Square room length (meter) 1 30 mx 30 m Omni+DFE . 1 Pout Omni Sector . 01. 001 . 0001 Sec+DFE 0 10 20 30 Rb (Mbps) 40 50 [G. Yang and K. Pahlavan, "Comparative Performance Evaluation of Sector Antenna and DFE Systems in Indoor Radio Channels, " Proc. of ICC '92] 8 C 32810. 122 -Cimini-7/98

CHANNEL ACCESS ISSUES • Multiple Access • Random Access • Frequency Reuse 8 C 32810. 125 -Cimini-7/98

MULTIPLE ACCESS TECHNIQUES • Frequency Division (FDMA) • Time Division (TDMA) • Code Division (CDMA) • Hybrid Approaches 7 C 29822. 033 -Cimini-9/97

FDMA The total system bandwidth is divided into channels which are allocated to the different users. Code Space Time Frequency 7 C 29822. 030 -Cimini-9/97

TDMA Time is divided into slots which are allocated to the different users. Code Space Time Frequency 7 C 29822. 031 -Cimini-9/97

CDMA Time and bandwidth are used simultaneously by different users, modulated by orthogonal or semiorthogonal codes (e. g. spread spectrum). Code Space Time Frequency 7 C 29822. 032 -Cimini-9/97

IMPLICATIONS FOR HIGH-SPEED WIRELESS DATA • Perform well with continuous stream traffic but inefficient for bursty traffic • Complexity Frequency Division < Time Division < Code Division • Multiple Data Rates – multiple frequency bands – multiple timeslots – multiple codes 7 C 29822. 034 -Cimini-9/97

RANDOM ACCESS TECHNIQUES • ALOHA • Carrier-Sense Techniques • Reservation Protocols • Implication for High-Speed Wireless Data 7 C 29822. 038 -Cimini-9/97

ALOHA • Data is packetized. • Retransmission is required when packets collide. • Pure ALOHA – send packet whenever data is available – a collision occurs for any partial overlap of packets • Slotted ALOHA S (Throughput per Packet Time) – send packets during predefined timeslots – avoids partial overlap of packets . 40 Slotted Aloha . 30. 20 Pure Aloha . 10 0 0. 5 1. 0 1. 5 2. 0 3. 0 G (Attempts per Packet TIme) • Comments – inefficient for heavily loaded systems – capture effect improves efficiency – combining SS with ALOHA reduces collisions 8 C 32810. 39 -Cimini-7/98

CARRIER-SENSE TECHNIQUES • Channel is sensed before transmission to determine if it is occupied. • More efficient than ALOHA Þ fewer retransmissions • Carrier sensing is often combined with collision detection in wired networks (e. g. , Ethernet). Þ not possible in a radio environment Busy Tone Wired Network Wireless Network • Collision avoidance is used in current wireless LANs. (Wave. LAN, IEEE 802. 11, Spectral Etiquette) 8 C 32810. 40 -Cimini-7/98

RESERVATION PROTOCOLS • Demand–Based Assignment – a common reservation channel is used to assign bandwidth on demand – reservation channel requires extra bandwidth – very efficient if overhead traffic is a small percentage of the message traffic • Packet Reservation Multiple Access (PRMA) – similar to reservation ALOHA – uses a slotted channel structure – all unreserved slots are open for contention – a successful transmission in an unreserved slot effectively reserves that slot for future transmissions 7 C 29822. 041 -Cimini-9/97

EXAMPLES • ARDIS – slotted CSMA • RAM Mobile Data – slotted CSMA • CDPD – DSMA/CD - Digital Sense Multiple Access – collisions detected at receiver and transmitted back • Wave. LAN – CSMA/CA 8 C 32810. 126 -Cimini-7/98

IMPLICATIONS FOR HIGH SPEED WIRELESS DATA • Retransmissions are power and spectrally inefficient. • ALOHA cannot satisfy high-speed data throughput requirements. • Reservation protocols are also ineffective for short messaging. • Delay constraints impose throughput limitations. 7 C 29822. 042 -Cimini-9/97

FREQUENCY REUSE BASE STATION • Frequencies (or time slots or codes) are reused at spatially-separated locations. • Introduces interference Þ system capacity is interference-limited. • Mainly designed for circuit-switched communications • Base stations perform centralized control functions. (call setup, handoff, routing, etc. ) 8 C 32810. 43 -Cimini-7/98

DESIGN CONSIDERATIONS • Reuse Distance (D) – distance between cells using the same frequency, time slot, or code – smaller reuse distance packs more users into a given area, but also increases their co-channel interference • Cell Radius – decreasing the cell size increases system capacity, but complicates the network functions of handoff and routing 8 C 32810. 44 -Cimini-7/98

CHANNEL ASSIGNMENT • Fixed Channel Assignment (FCA) – each cell is assigned a fixed number of channels – channels used for both handoff and new calls • Reservation Channels with FCA – each cell reserves some channels for hand off calls • Channel Borrowing – a cell may borrow free channels from neighboring cells • Dynamic Channel Assignment 7 C 29822. 67 -Cimini-9/97

METHODS TO IMPROVE SPECTRUM UTILIZATION • Interference Averaging (CDMA) • Interference Reduction (power adaption, sectorization) • Interference Cancellation (smart antennas, multiuser detection) • Interference Avoidance (dynamic resource allocation) 8 C 32810. 45 -Cimini-7/98

Ad-Hoc Networks • • • Each node generates independent data. Source-destination pairs are chosen at random. Routing can be multihop. Topology is dynamic Generally a fully connected network with different link SNRs • Can allocate resources dynamically (rate, power, BW, routes, …)

NETWORK ISSUES • Network Architectures • Mobility Management • Network Reliability • Internetworking • Security 8 C 32810. 53 -Cimini-7/98

NETWORK ARCHITECTURES • Hierarchical/Tree • Star • Ad-Hoc • Implications for High-Speed Wireless Data – single hop versus multiple hops – static versus dynamic topology – single points of failure 7 C 29822. 054 -Cimini-9/97

NETWORK CONTROL • Centralized – RAM Mobile Data – CDPD – Altair • Distributed/Peer-to-Peer – Wave. LAN • Implications for High-Speed Wireless Data – less channel estimation required with centralized control Þ increases efficiency of packet transmission – centralized control provides more efficient resource management with setup-time overhead – an extensive infrastructure is not required for distributed control 7 C 29822. 055 -Cimini-9/97

MOBILITY MANAGEMENT • Location Management – identification and authentication – home and visitor location data bases (cellular) – discovery and registration (Mobile IP) • Routing – fixed data bases (connection-oriented) – Mobile IP (connectionless) – tree (virtual connection) overhead and delay impact throughput suboptimal (triangle) routing Þ delay inefficiency and higher congestion • Handoff – transmissions may be delayed or dropped Þ impacts higher layer protocols – multi-homing Þ inefficient use of resources 8 C 32810. 131 -Cimini-7/98

NETWORK RELIABILITY • End-to-End connection is composed of many wireless/wired hops. – widely varying data rates – high BERs on some/all hops – large, varying latencies – user mobility causes hop characteristics to vary • Problem with reliability protocols like TCP. – wireless losses mistaken for congestion – bulk losses cause timeouts – large round-trip time variances and asymmetric channels 8 C 32810. 58 -Cimini-7/98

APPROACHES TO NETWORK RELIABILITY • Local (link-layer) solutions – Forward error correction does not work well in fading – ARQ introduces large latency • End-to-end solutions – Difficult to distinguish if packet loss due to congestion or link quality – Difficult to design for changing hop characteristics End-to-end performance guarantees are difficult to make • Potential solutions – Hierarchical/layered coding of voice/video/images – Different Quality-of-Service classes – Application awareness – Local solution with end-to-end awareness Requires interaction between all layers 8 C 32810. 58 a-Cimini-7/98

QUALITY OF SERVICE (Qo. S) • Traffic dependent performance metrics required for type of data transmitted – bandwidth – latency – likelihood of packet (message) loss • Categories – guaranteed – predictive – best effort • Implications for high speed wireless data – Qo. S performance generally based on switched, fiber-optic, wired networks – wireless links have high Pb and high latency due to link layer retransmission and unpredictable link bandwidths – Qo. S guarantees and predictions are difficult to make for wireless networks Þ it is not clear that the best effort is good enough for most applications 7 C 29822. 059 -Cimini-9/97

INTERNETWORKING • TCP/IP – Compatible with existing wired networks – Works well over large range of wired subnet performance – TCP has problems operating over wireless links • Wireless ATM – ATM is emerging standard for multimedia transmission over wired networks – ATM protocol based on links with 10 -10 BER and Mbps/Gbps data rates • high overhead in packet structure • QOS guarantees – Not clear that ATM protocol can be modified for wireless links 7 C 29822. 68 -Cimini-9/97

STANDARDS AND FUTURE SYSTEMS • Bluetooth • Wireless LANs • High-Speed Digital Cellular (3 G) • 4 G Cellular • Wireless "Cable" – Multichannel Multipoint Distribution Service (2. 2 GHz) – Local Multipoint Distribution Service (28 GHz) • Satellite Networks - Iridium, Globalstar, Others • Home. RF 8 C 32810. 61 -Cimini-7/98

BLUETOOTH • Cable replacement RF technology • Short range (10 meters) • 2. 4 GHz band • 1 Data (700 Kbps) and 3 Voice channels • Supported by over 200 telecommunications and computer companies 8 C 32810. 61 -Cimini-7/98

802. 11 Wireless LANs • 802. 11 b: standard for 2. 4 GHz ISM band • Frequency hopped spread spectrum • 1. 6 Mbps data rates, 500 foot range • Star or peer-to-peer architecture • 802. 11 a extends rates to 10 -70 Mbps • Extensions trying to add Qo. S 8 C 32810. 63 b-Cimini-7/98

HIPERLAN • Types 1 -4 for different user types - Frequency bands: 5. 15 -5. 3 GHz, 17. 117. 3 GHz • Type 1 - 5. 15 -5. 3 GHz band - 23 Mbps, 20 MHz Channels - 150 foot range (local access only) - Protocol support similar to 802. 11 - Peer to peer architecture - ALOHA channel access • Types 2 -3 - Wireless ATM - Local access and wide area services - Standard under development - Two components: access and mobility support 8 C 32810. 63 a-Cimini-7/98

HIGH-SPEED DIGITAL CELLULAR • North American Digital Cellular – CDMA (IS-95) enhancements – TDMA (IS-136) enhancements – IS-136+ Þ 32 -64 kbps – IS-136 HS Þ 384 kbps • GSM – General Packet Radio System (GPRS) – Enhanced Data Rates for GSM Evolution (EDGE) 8 C 32810. 62 -Cimini-7/98

EDGE • Evolution of GSM / GPRS • ETSI standardization as GSM evolution chosen for data services for IS 136 HS 8 C 32810. 137 ppt-Cimini-7/98 • Higher-level modulation (adaptive) • 200 k. Hz carrier spacing • Up to 384 kbps in 200 k. Hz

WIDEBAND CDMA (3 G) • The W-CDMA concept: – 4. 096 Mcps Direct Sequence CDMA – Variable spreading and multicode operation – Coherent in both up-and downlink = Codes with different spreading, giving 8 -500 kbps . . P f 4. 4 -5 MHz t High rate multicode user Variable rate users 8 C 32810. 138 ppt-Cimini-7/98 10 ms frame

W-CDMA KEY TECHNICAL FEATURES • High bit-rate services require wideband • Flexibility for different services • Optimized for packet data transfer • Capacity and coverage gain from frequency diversity • Built in support for – adaptive antenna arrays – multi-user detection – hierarchical cell structures – transmitter diversity • Low infrastructure cost (many users/ transceiver) • BS synchronization not required 8 C 32810. 139 ppt-Cimini-7/98

SUMMARY • The desire for mobility coupled with the demand for Internet and multimedia services indicate a bright future for wireless data. • Current products and services have unsatisfactory performance for high-speed wireless data applications. • The inherent limitations of the radio channel can be significantly reduced using signal processing and architectural techniques, at the expense of cost and complexity. • The network-level design must take into account the physical layer limitations of the wireless channel, as well as the impact of user mobility. 8 C 32810. 65 -Cimini-7/98